How Pattern Recognition Will Solve Problems — Part 2: The Sporadic Pattern
How to solve problems that occur (seemingly) randomly?
Fascinating problem this one…?
An ATM in a bank branch was causing some serious problems.?
The general pattern was:?
It would be working fine, but then all of a sudden, it would start rebooting… multiple times a day, sometimes four or five times, sometimes even more and then for a couple of days, it didn't actually reboot at all. But it did seem to reboot most days at least once or twice.?
So…the engineers started replacing the parts in the ATM. They initially thought it was a power supply issue. Then, they thought it was some sort of motherboard issue. Some other parts started getting replaced, and they kept replacing parts.?
But the ATM just kept rebooting.
When it comes to assessing problems or issues, as we discussed in Part 1, pattern recognition is one of the most valuable tools at your disposal.?Click this link below to read part 1 if you missed it.
In fact, it’s something that helps determine the vast majority of causes.?
But remember, sporadic patterns are much more challenging to resolve.
Now, here’s something really important. If there’s a regular (periodic) or constant (continuous) pattern, you can often find the problem within the ‘system’ that governs whatever it is you’re investigating.?
Sporadic patterns come and go… by their very nature, you can’t predict them. (Otherwise, they’d be continuous or periodic patterns.)
So, they require ‘out-of-the-box’ thinking. The solution is usually not obvious or evident within the system.?
We see many teams wasting an incredible amount of time diagnosing and examining things that they can control. They'll examine the configurations on their applications, firewalls, or specific network devices.
Or if it's a physical thing like a car or something else, they'll be looking at everything they can control. You know it's generally not going to be something you can control, as the very nature of sporadic patterns, in 98% of cases, is usually caused by elements that are out of your control.
领英推荐
Let’s take a look at an example:?
In one scenario we worked with, a user had developed their own poorly built manual query. And when they ran this query on the network, it sucked all the resources out of the database server. This caused what was seen as the network to slow down whenever they decided to run it.?
Thus, the pattern you’d observe would look completely random.?
So, where would you look??
Remember, what you can control usually is not the problem. Think about what could impact your application… what does the app interface with? Does it travel over certain links? Could someone be doing something to the link? What factors external to the App or thing could be influencing it?
So, going back to our ATM example.
Once they had exhausted the investigation of replacing parts in the machine, they determined that it must be a systemic issue with the ATM itself. So, they removed the ATM, and they took an extra ATM out of a branch that wasn't getting used as much.?
They took it out of that branch and put it into this branch in the same spot. Now, that ATM was working fine in the previous branch, but when they plugged it into this branch, it started rebooting.
At this point, they were making progress, albeit very slow, because they had effectively ruled out the ATM itself.?
So, we prompted the team to consider external factors, as the issue may not stem solely from the ATM itself.
They decided to investigate externally around the ATM. Specifically behind the wall that this ATM was against, they found that the office next door had been vacant for quite some time.
But recently, a specialist doctor had moved into that office. When they went and looked in the office, the doctor had installed a specialist imaging machine directly behind the ATM. Now, whenever they turned this machine on, which is basically this big magnet, it caused an electromagnetic type of interference.
This caused the ATM to reboot. So, whenever this machine was turned on throughout the day, the ATM would reboot.?
Moral of the story: when you get a truly sporadic issue like this, look outside the box and think differently. Think about what external factors, that you can’t control, could be affecting you.
For more information, contact me directly or learn more at Home (thinkingdimensions.com.au)
Simplifying The Complexity That Is Change // Navigating Through Constant and Unprecedented Change With Ease // Organizational Change, Leadership Capability Uplift, Workforce Resilience, High Performing Distributed Teams
5 个月Great insights. Thanks Andrew